Flexible, Mechanically Durable Aerogel ... - ACS Publications

Dec 24, 2015 - for automated oil capture and recovery: Cabot Thermal Wrap (TW) and Aspen ... Macondo Blowout disaster (April 2010), released approx-...
0 downloads 0 Views 3MB Size
Research Article www.acsami.org

Flexible, Mechanically Durable Aerogel Composites for Oil Capture and Recovery Osman Karatum,† Stephen A. Steiner III,‡ Justin S. Griffin,‡ Wenbo Shi,§ and Desiree L. Plata*,†,§ †

Department of Civil and Environmental Engineering, Hudson Hall, Duke University, Durham, North Carolina 27707, United States Aerogel Technologies, LLC, Boston, Massachusetts 02127, United States § Department of Chemical and Environmental Engineering, Mason Laboratory, Yale University, New Haven, Connecticut 06511, United States ‡

S Supporting Information *

ABSTRACT: More than 30 years separate the two largest oil spills in North American history (the Ixtoc I and Macondo well blowouts), yet the responses to both disasters were nearly identical in spite of advanced material innovation during the same time period. Novel, mechanically durable sorbents could enable (a) sorbent use in the open ocean, (b) automated deployment to minimize workforce exposure to toxic chemicals, and (c) mechanical recovery of spilled oils. Here, we explore the use of two mechanically durable, low-density (0.1−0.2 g cm−3), highly porous (85−99% porosity), hydrophobic (water contact angles >120°), flexible aerogel composite blankets as sorbent materials for automated oil capture and recovery: Cabot Thermal Wrap (TW) and Aspen Aerogels Spaceloft (SL). Uptake of crude oils (Iraq and Sweet Bryan Mound oils) was 8.0 ± 0.1 and 6.5 ± 0.3 g g−1 for SL and 14.0 ± 0.1 and 12.2 ± 0.1 g g−1 for TW, respectively, nearly twice as high as similar polyurethane- and polypropylene-based devices. Compound-specific uptake experiments and discrimination against water uptake suggested an adsorption-influenced sorption mechanism. Consistent with that mechanism, chemical extraction oil recoveries were 95 ± 2 (SL) and 90 ± 2% (TW), but this is an undesirable extraction route in decentralized oil cleanup efforts. In contrast, mechanical extraction routes are favorable, and a modest compression force (38 N) yielded 44.7 ± 0.5% initially to 42.0 ± 0.4% over 10 reuse cycles for SL and initially 55.0 ± 0.1% for TW, degrading to 30.0 ± 0.2% by the end of 10 cycles. The mechanical integrity of SL deteriorated substantially (800 ± 200 to 80 ± 30 kPa), whereas TW was more robust (380 ± 80 to 700 ± 100 kPa) over 10 uptake-and-compression extraction cycles. KEYWORDS: aerogel blanket, oil remediation, oil sorbent, mechanical extraction, polyurethane foam



construction materials4 and the challenge of maintaining a perfectly intact or vertical physical barrier during high currents and winds. This allows oil to flow over and underneath the boom, ultimately impacting sediments and shore. Intense oil contamination in some parts of Louisiana marshes caused death of ecologically valuable plants.5 Industrial recovery responses favor the use of skimmers (i.e., skimmer barrels), but these can only be deployed close to vessel and have optimal performance when the surface oil layer is thick.6,7 Skimmers cannot be applied in deep sea (i.e., for well blowouts) or in conjunction with dispersants, as dispersants thin and break up the oil layer to promote aerobic biodegradation. Sorbent pads are used frequently to remove remaining oils from shore and at sea, but these never remove oil comprehensively, and they require manual deployment and recovery in current configurations. The most commonly used industrial sorbent materials are polypropylene (PP) and polyurethane foams (PUF),8,9 as they are affordable, lightweight, and readily deployable. However, they suffer from being fairly nonspecific sorbents, taking up

INTRODUCTION Increasing energy demand, due in part to the growing global population, has led to continuous rise in global consumption of crude oil.1 In association with this demand, more than 7.3 million barrels of oil enter the natural aquatic environments from municipal and industrial sources, marine transport, natural oil seeps, and accidents.2 The largest oil spill in U.S. history, the Macondo Blowout disaster (April 2010), released approximately 4.9 million barrels of Southern Louisiana crude oil into Gulf of Mexico.3 The second largest oil spill in the Gulf of Mexico, the Ixtoc I well blowout (3 million barrels), occurred in June 1979, and in spite of over 30 years separating the two and enormous material innovation during that time, the responses to the disasters were strikingly similar (Supporting Information (SI), Table S1). Commonly, physical (e.g., containment booms, skimmers, and sorbents), chemical (e.g., dispersion, in situ burning and solidifiers), and biological (e.g., biostimulation, natural attenuation, and bioaugmentation) methods are applied for the removal of oil from surface or shore. Containment booms are the most frequently used piece of equipment for preventing spilled oil from spreading, especially into biologically sensitive areas and coastal regions. However, oil booms are known to be unstable due to both the lack of robust © XXXX American Chemical Society

Received: September 8, 2015 Accepted: December 12, 2015

A

DOI: 10.1021/acsami.5b08439 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

discrimination against water uptake. To meet these important needs, we evaluate the use of two hydrophobic, lightweight, and mechanically durable flexible aerogel fabrics for oil spill removal and recovery: Cabot Thermal Wrap (TW) and Aspen Aerogels Spaceloft (SL). Currently, these materials are used primarily for insulation applications, including apparel, building, and industrial insulation, and no commercial applications rely on their unique surface chemistry. Thus, the use of TW and SL as mechanically robust oil sorbents presents a truly novel application of the materials and a transformative approach to oil remediation and recovery.

water along with oil. This has several consequences, including reduced oil uptake capacity and eliminating the possibility of economically competitive oil recapture. In particular, if oil reclamation is sought, additional approaches such as centrifugation, filtration, or gravitational settling must be used to remove the water.10 More often, PP and PUF are discarded without oil recovery and ultimately incinerated. Some have proposed the use of waste materials, such as human hair11 and old PUF furniture as sorbents, but these have clear disadvantages with respect to a readily available material supply as well as challenges in systematic mechanical recovery (i.e., scooping oiled hair out of the ocean is not straightforward). Ideally, sorbents should be reusable, provide efficient and specific recovery of oil, require little human intervention for deployment and retrieval, and have some mechanical integrity. In particular, mechanically durable sorbents would open up a world of previous unrealized applications, such as (a) booms able to withstand ocean towing forces, (b) automatically deployed and recovered sorbents (i.e., using robots that skim the ocean surface or deploy and recover sorbent mats in coastal areas), (c) sorbent use in high wind and waves, and ultimately, (d) all of these could be followed by mechanical recovery of the collected oil. Considering the achievements in materials science over the past several decades, advanced composite materials could meet these needs. For example, aerogels are sol-gel-derived, highly porous, lightweight solid materials with some of the highest surface areas per mass of all known materials.12 Recent advances in aerogel fabrication and functionalization (i.e., moving beyond classic silica aerogels) are enabling multiple tunable properties, including mechanical strength, flexibility, stability in water, hydrophobicity, and conductivity. These have given rise to a variety of environmental applications, such as removal of miscible and immiscible organic solvents in water;13,14 volatile organic compound (VOC) vapors;15 toxic organic compounds;16 and oil.17,18 For example, perfluoroalkylated (i.e., −CF2 and −CF3 groups) hydrophobic aerogels remove oil masses up to 14 times their dry weights19 when in powder form. While silica aerogels exhibit efficient removal of target chemicals, their brittle, fragile, and nonelastic characteristics limit their practical applications.20,21 Some cellulose aerogels exhibit enhanced physical integrity and have been demonstrated for oil extraction but degraded after a single mechanical extraction (i.e., squeezing for oil recovery), and their sorption capacity significantly diminished during the second cycle.22 Carbon aerogels with fibrous morphologies (i.e., cylindrical shaped and made of carbon; approximately 5 μm in diameter) manufactured from different carbon sources such as cotton fiber and bacterial cellulose had similar, albeit slightly improved, mechanical recycle performance. Distillation and combustion recharge approaches were also tested for these materials;23 however, these “extraction” methods are energy consumptive and present clear flammability hazards as compared to mechanical compression recovery approaches. As a result, the large-scale applications of these latter methods are limited. Thus, while sorbents themselves are a mature technology, there is a need for enhanced sorbent materials that (a) have a continuous geometry to protect sensitive areas when booms cannot; (b) are mechanically durable, enabling ocean towing, automated deployment with minimal human intervention, and mechanical extraction to recover oil; and (c) provide a costefficient material that has a high selectively for oil and



EXPERIMENTAL APPROACH

Materials. Thermal Wrap (TW) (6.0 mm thickness, Cabot Corporation, Boston, MA) and Spaceloft (SL) (10 mm thickness, Aspen Aerogels, Northborough, MA) are flexible fiber-reinforced silica aerogel composite blankets. Both materials utilize hydrophobic trimethylsilylated silica aerogel. The trimethylsilyl functionalization imparts hydrophobicity and aqueous stability to the silica aerogel that is natively absent. Thermal Wrap is a composite of trimethylsilylated silica aerogel particles with bicomponent polyester fibers that binds the particles into a low-dust flexible blanket. Spaceloft is a composite of a lofty fibrous batting of polyester and glass fibers infiltrated with trimethylsilylated silica aerogel and magnesium hydroxide. Both products are commercially available and produced on large scales. All solvents (dichloromethane (DCM), methanol, acetone, and hexanes) were GC-grade (EMD Chemicals). Seawater was prepared in accordance with ASTM D1141-98 using sea salt from Lake Products LLC (Florissant, MO). Briefly, this method involves dissolving 41.9 g of commercially produced sea salt in 1 L of water. Iraq and Sweet Bryan Mound (SBM) crude oils were obtained from ONTA, Inc. (Ontario Canada). A competing material, PUF, which is used in the majority of oil spill responses in near-shore applications, was purchased from Foam Factory, Inc. and tested alongside the TW and SL aerogel composites. Material Characterization. To characterize the morphology of the aerogel composites, we coated test swatches with a 25 nm chromium layer via a Denton chromium sputtering machine and imaged by a Hitachi scanning electron microscope (SEM) SU-70. Surface area measurements were performed using a TriStar II 3020 porosimeter (Micromeritics) with N2 as an adsorbing gas. Surface area was derived from the adsorption isotherm using Brunauer, Emmett, and Teller (BET) theory. Pore size distribution was derived from the desorption isotherm using Barret−Joyner−Halenda (BJH) theory. Skeletal density of the materials was determined using an AccuPyc II helium pycnometer with 10 purges and 200 measurements for each sample, using a 10-mL sample chamber. Then, porosity was determined by taking a sample of known volume, subtracting the skeletal volume, and expressing the pore volume as a percentage of total volume. Contact Angle Measurements. Water contact angles were measured via the sessile drop method using a Kruss DSA-20 instrument and an average of 10 different spot measurements. Due to the fibrous nature of the material, the baseline determinations were made using the tangent of the bottom of the water droplet, which approximated an average through the opaque surface of the aerogel composite mat (see SI Figure S1 for example). Sorption Capacity Experiments. In the absence of water, preweighed TW and SL aerogel swatches (20 × 20 × 6 mm and 20 × 20 × 10 mm, respectively) were completely immersed in a solvent or oil inside an open, precombusted glass beaker using a precombusted Pasteur pipet to submerge the aerogel gently. After each 2 min immersion interval over a 30 min duration, aerogels were removed and weighed. This experiment was performed in triplicate and sorption capacity was calculated as follows:

sorption capacity (g/g) = B

Mf − M0 M0

(1)

DOI: 10.1021/acsami.5b08439 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces Table 1. Physical Characteristics of Blanket Aerogels and a Comparative Oil Sorbent Thermal Wrap (TW) skeletal density (g mL−1) bulk density (g mL−1) bulk density after oil uptake (g mL−1) porosity (%) BET SA (m2 g−1) tensile strength (kPa) water contact angle (°) composition manufacturing method

Spaceloft (SL)

PUF

1.56 ± 0.01 0.12 0.37 ± 0.08

1.73 ± 0.02 0.24 0.63 ± 0.08

0.0234 0.12 ± 0.03

92.0 ± 0.1 407 380 ± 83 (n = 5) 127 ± 6 PE fiber with Trimethylsilylated silica hydrophobic silica attached to fibers with heating

86.0 ± 0.1 528 766 ± 244 (n = 6) 130 ± 9 glass fiber, PE fiber with triethylsilylated silica supercritically drying a whole composite (i.e., fibrous batting infiltrated with sol gel)

92 ± 7 (n = 3) 6432 polyurethane35 controlled expansion35

PE, polyester; PUF, polyurethane foam, a common oil sorbent shown for comparison. Unless otherwise noted, all measurements were collected as part of this study.

Figure 1. Optical interrogation of aerogel fabrics’ gross and fine structure. Image of a dyed-green water droplet and Iraq oil on the surface of (a) TW (6 mm thick) and (d) SL (10 mm thick). Insets are images from the WCA measurements. Note that the TW used is thinner than SL. SEM images of the surfaces of (b and c) TW and (e and f) SL, where aerogel particles are visible on the fibers in the high-resolution images (c and f).

sorption capacity (mL oil/ mL aerogel pore space) M f − Mo ρ 1 = × a × Mo ρs Φ

Chemical Extraction. Aerogel samples were spiked with four perdeuterated recovery standards (2-methylnaphthalene-d10, anthracene-d10, p-terphenyl-d14, benzo[e]pyrene-d12, and recoveries ranged from 60% to 95%). Accelerated solvent extraction (ASE Thermo Dionex) at high pressure and temperature (1000 psi, 100 °C) eroded the aerogel structure to a point at which fine particles could not be well separated from the oil extract, and so gentler Soxhlet extraction was employed at 60 °C for 24 h in a 90:10 mixture of dichlormethane:methanol (DCM:MeOH).24 Extracts were concentrated by rotary evaporation and spiked with six perdeuterated injection standards (naphthalene-d8, acenanaphthalene-d10, phenanthrene-d10, chrysene-d12, perylene-d12, and dibenz[a,h]anthracene-d14). Diesel range organic compounds and target polycyclic aromatic hydrocarbons (PAHs) were analyzed by using an Agilent 7890 gas chromatograph (GC) equipped with an Agilent 5975 quadrupole mass spectrometer (MS) and an HP-5 column (30 m × 0.32 mm ID × 0.25 μm film thickness). Compound identities were confirmed using mass spectral library searches (National Institute of Standards and Technology (NIST) Library) and retention times from standard compounds. The total dry lipid content of each extract (i.e., the total lipid extract (TLE)) was determined gravimetrically by transferring 50−250 μL aliquots into a preweighed aluminum dish and allowing them to dry until a constant weight was observed. Note that the TLE of the neat oil prior to any extraction or recovery was 75 and 70% w/w for Iraq Oil and SBM Oil, respectively. Mechanical Extraction. Using a Palmgren drill press vise device with 38 N of pressure, oil-saturated aerogels were compressed while

(2)

where Mo is initial dry absorbent (aerogel) weight, Mf is weight of sorbent samples at the end of tests, ρs is the density of fluid (i.e., solvent or oil), ρa is the bulk density of the aerogel, V is the volume, and ϕ is the porosity of aerogels. Note that porosity (92 ± 0.1% for TW, 86 ± 0.1% for SL; Table 1) is applied for the calculation of sorption capacity based on the volume. To test the effect of seawater on the sorption capacity of TW and SL aerogels, 50 mL of artificial seawater was added to each beaker along with a sufficient amount of oil. In this case, aerogels were allowed to float on the surface of the seawater as they would normally for 30 s. Note, oil uptake was rapid and seemed to stabilize after 30 s based on an optical observation of the darkness of the remaining oil slick. Then, the beaker was mixed to simulate a turbulent environment. The final mass of aerogel was measured after 15 min of uptake and performed in triplicate. Blanks (i.e., aerogels in seawater without any added oils) were measured in triplicate also. Recovery and Reusability. In order to assess the potential for oil recovery from aerogel sorbent fabrics, we compared chemical and mechanical extraction approaches. Typically, the former is considered to be robust and efficient (i.e., high recovery), but is not amenable for field deployment. In contrast, mechanical extraction can have highly variable recoveries, but is inherently field amenable. C

DOI: 10.1021/acsami.5b08439 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces suspended over a precombusted, preweighed aluminum pan for 30−60 s, until no further oil extruded from the compressed aerogel. The resultant oil was weighed immediately, and this process was repeated for 10 cycles. Oil recovery was calculated via TLE mass. Tensile Strength Testing. Material samples for tensile strength testing were cut from bulk blanket material into pieces 25 × 150 mm as specified in the cut strip test (applicable to nonwoven fabrics) described in ASTM D5035.25 Initially, each material was measured in replicate (n = 6 for SL, 5 for TW, and 3 for PUF) to determine the relative standard deviation of the tensile strength for a given material, reflecting both analytical variability and material heterogeneity. To probe the changes in the tensile strength as a result of oil uptake and extraction, we then placed individual samples in a saltwater bath with ∼10 g soybean oil for 10 min. (See SI Section S3 for justification of choice of soybean oil as a surrogate for crude oil in these experiments). After uptake, the sample was placed between two stainless steel plates, to which a load of 10 kN was applied for 60 s to expel the sorbed oil. This process was repeated until the desired cycle number was reached. Sample dimensions were measured before loading the sample into an Instron 3366, held by side-action screw type grips set 75 mm apart. Tensile load was applied to the sample at a constant rate of extension (50 mm s−1; lower than the 300 mm s−1 specified in ASTM D5035) until the material failed. Finally, stress was calculated from the maximum load sustained by the sample and the pretest dimensions.

where positive Wi corresponds to a spontaneous immersion process. For oil, contact angles of zero indicate spontaneous immersion of the solid aerogel in the liquid oil. In contrast, for water, the work of immersion is negative and immersion of the solid aerogel in water is nonspontaneous and disfavored. Macroscopically and microscopically, the structures of TW and SL appear to be similar (Figure 1, Table 1). In particular, the skeletal densities, internal surface areas, and porosities of the two materials are similar. They differ with respect to both bulk density (0.12 and 0.24 g mL−1 for TW and SL, respectively) and mechanical strength (380 ± 80 and 800 ± 200 kPa tensile strength for TW and SL, respectively). In addition, there are differences in the manufacture of each material that could give rise to unique mechanical extraction and chemical uptake performance. SL is made by infilitrating a fibrous batting (fiber glass and polyester fibers) with wet gel and then supercritically drying the whole composite. In contrast, TW is made by mixing polyester fibers with silica aerogel particles and heating the mixture until an outer “skin” flows and adheres the aerogel particles to one another and to the fibers. Note that the melting temperature of the outer skin is distinct from the internal components of TW. As a result of compositional differences between TW and SL, we postulated they could exhibit separate uptake ability toward solvents, oils, and water with time and reuse cycles. Solvent and Oil Uptake Capacities. All tested organic solvents and oils reached equilibrium in very short time (